论文部分内容阅读
图像边缘检测是激光雷达图像处理的关键环节,传统激光雷达图像边缘检测装置在低照度条件下,其判决阈值会受到较大影响,导致检测效果较差。提出性能优异的激光雷达图像边缘检测器,检测器的硬件系统由控制中心模块、图像采集模块、图像配置模块、复原模块、边缘检测模块、VGA控制模块及图像存储模块组成;基于粒子群优化算法原理,给出了激光雷达边缘检测的软件工作流程,初始化随机粒子群,并进行迭代寻优。为避免陷入局部最优,对算子进行变异化处理,用激光雷达图像的像素代表粒子群中的元素,提升全局寻优的能力,实现激光雷达图像的边缘检测。该检测器具有边缘检测时间少,边缘检测精度的优点。
Image edge detection is the key point of LIDAR image processing. Under the condition of low illumination, the edge detection device of traditional lidar will be greatly affected, which leads to the poor detection effect. The hardware system consists of control center module, image acquisition module, image configuration module, recovery module, edge detection module, VGA control module and image storage module. Based on the particle swarm optimization algorithm Principle, the software workflow of laser radar edge detection is given, the stochastic particle swarm is initialized, and the optimization is performed iteratively. In order to avoid being trapped in local optima, the operator of the algorithm is transformed and the pixels in the Lidar image are used to represent the elements in the particle swarm optimization to enhance the ability of global optimization. The edge detection of Lidar images is achieved. The detector has the advantages of less edge detection time and edge detection accuracy.